cuPQC-BigInt

cuPQC-BigInt provides fixed-width, unsigned multi-precision integer arithmetic that runs directly inside CUDA kernels—single-thread or warp-cooperative (TPI) execution, modular and Montgomery arithmetic, and the wide (double-width) products needed to build modular-exponentiation and other big-number workloads on the GPU.

• Features — Execution modes, bit widths, thread-per-instance (TPI) configurations, supported operations, and error handling.

• Usage Guide — Descriptors, operators, thread and warp execution, arithmetic, division, modular and Montgomery arithmetic, compile and link.

• API Reference — Operators, types, and device functions.